Abstract Banks have to increase their capital ratios to meet the new regulation requirements. This can be achieved through raising new equity capital or by reducing bank assets. Banks often opt for the second channel due to the higher private costs of equity issuance. This, however, is associated with high public costs. In this background document we discuss the welfare costs of undercapitalized banks and review the empirical literature. Undercapitalized banks are bad for the economy mainly for two reasons. The first is misallocation of credit resulting from banks gambling for resurrection and reluctance to recognize losses. The second is contraction of credit supply driven by banks reducing their assets to increase their capital ratios. In this respect, we describe the experience in previous financial crises. Further, we analyze the potential private costs of bank recapitalization through equity issuance from a theoretical and empirical perspective. We perform an event study on bank equity issuance in the EU and US for the period 2007-2013. The results show a significant negative effect of SEOs on bank stock prices. However, this effect is on average not extreme and is comparable with other events. The negative effect on stock prices is partially compensated by a significant reduction of banks’ funding costs proxied by CDS spreads. The analysis concludes with a discussion of the appropriate policy options to enhance bank recapitalization. 3 Table of contents 1 Introduction ..................................................................................................................................................... 5 2 Costs of undercapitalization: empirical evidence ............................................................................ 5 2.1 Misallocation of credit ................................................................................................................................. 5 2.2 Contraction of credit supply ...................................................................................................................... 7 2.3 Case Studies ...................................................................................................................................................... 9 3 Theory: is bank recapitalization expensive? ................................................................................... 17 4 Empirics: stock market reactions to recapitalization .................................................................. 19 4.1 Introduction .................................................................................................................................................. 19 4.2 Event study methodology ........................................................................................................................ 20 4.3 Data and descriptive statistics............................................................................................................... 22 Note: Mean values on top and standard deviations in parenthesis below. ..................................... 24 4.4 Results ............................................................................................................................................................. 24 4.5 Robustness..................................................................................................................................................... 30 4.6 Conclusions from the event study ........................................................................................................ 31 5 Policy options ............................................................................................................................................... 31 Literature .................................................................................................................................................................... 33 4 1 Introduction In response to stricter regulation, banks have to raise their capital ratios. In principle, they can achieve this through raising new equity capital, but also by reducing their assets. Whereas the former may be the intended response by regulators, banks might be reluctant to issue new equity and rely more on the second channel. The motivation for this choice is that banks consider equity issuance as more expensive. The alternative however is associated with high public costs. The purpose of this study is to provide insights on several important questions related to the private and public costs of bank recapitalization. First, we consider the welfare costs of undercapitalized banks and survey the empirical literature. Delay in recapitalization and a slow increase in equity ratios through a shrinking of assets can have adverse social welfare effects. In this respect, we describe the experience in previous financial crisis. Second, we discuss the potential private costs of bank recapitalization from a theoretical perspective and empirical perspectives. The main empirical part of this paper consists of surveying the data on bank equity issuance in the period 2007-2013 in the US and the EU. We compare banks’ stock issuances in both regions, and test empirically whether they affected their market value and funding costs. We conclude with a discussion of policy options to take away possible obstacles to speedy recapitalization. 2 Costs of undercapitalization: empirical evidence Banks’ undercapitalization has substantial public costs. They are well documented from previous crises. This section outlines the main literature and summarizes empirical evidence from past crises. We can distinguish two types of social costs. One is that undercapitalization induces distorted decisions for banks, resulting in a misallocation of capital. The second is that if banks decide to increase their capital ratios in other ways - in particular by shrinking their assets - the associated credit constraints may impede the growth of healthy firms. 2.1 Misallocation of credit Low capitalization creates distortions in banks’ decisions due to risk shifting: banks that have a realistic chance of going bankrupt - and perhaps subsequently getting bailed out - do 5 not fully internalize the downside of investments they make. In case of failure, the debt holders, or the government, will bear part of the burden of the bad investment. In case of success, on the other hand, the gains on the risky investment may help the bank to recover from its troubles, to the benefit of shareholders and management. This “heads I win, tails you lose” strategy is therefore sometimes termed “gambling for resurrection”. Gambling for resurrection, in the form of misallocation of credit as a result of an undercapitalized banking system, has been extensively studied in the context of the Japan financial crisis (see e.g. Hoshi and Kashyap, 2013, for an overview). Troubled banks had perverse incentives to keep credit flowing to firms that were insolvent, in order to avoid the recognition of losses that would result from terminating lending relationships with those firms. This forbearance allowed the poorly capitalized banks to keep up appearances of sufficient capitalization, and postpone painful restructuring, in the hope that problems would be resolved with a recovery in the Japanese economy. Similar behaviour of gambling for resurrection was also observed in the US, during the Savings and Loans crisis in the 1980s, where the introduction of Prompt Corrective Action finally put a stop to that (see Benston and Kaufman, 1997).The effect of lending to insolvent firms (and inhibiting growth for stronger firms) is termed zombie lending and is a specific type of gambling for resurrection. Zombie lending has been extensively studied in the literature (see e.g. Hoshi and Kashyap, 2013, for references). In one important analysis, Peek and Rosengren (2005) study bank lending patterns in Japan in the 1990s. They find that Japanese banks provided subsidized loans in particular to weak or insolvent firms, and that this occurred more if the bank itself had a weak capitalization. They explain the result as a form of “balance sheet cosmetics”, a desire to hide losses in order to avoid forced recapitalization. Zombie lending not only had adverse effects for the banks involved, but also created negative spill-over effects on Japan’s economy. Caballero et al. (2008) analyze the effects on Japanese firm dynamics as a result of zombie lending. Keeping unprofitable firms alive worsens business opportunities for more efficient firms and keeps them from growing. The continued presence and subsidization of inefficient firms reduced efficient firms’ profitability, as well as their ability to attract funding because of reduced value of their collateral. The authors find that healthy firms in sectors with more zombie-lending invest less and create fewer jobs. In contrast to these Japanese studies, Drechsler et al. (2013) look at risk shifting by banks in Europe, in the current crisis. They study bank-level data on repo lending in the euro zone by the ECB, in the period 2007-2011. This lending was provided in order to provide liquidity support to banks, under collateral requirements that were less strict than those of the market. They find that more weakly capitalized banks borrow more, and also at riskier collateral, in particular sovereign bonds. In addition, these banks increased their holdings of this riskier collateral during the period. Drechsler et al. (2013) argue that at least part of the lending was not explained by illiquidity